CN105636062B - A kind of cognition wireless network that service-oriented moderately services transmission learning method - Google Patents
A kind of cognition wireless network that service-oriented moderately services transmission learning method Download PDFInfo
- Publication number
- CN105636062B CN105636062B CN201610046073.3A CN201610046073A CN105636062B CN 105636062 B CN105636062 B CN 105636062B CN 201610046073 A CN201610046073 A CN 201610046073A CN 105636062 B CN105636062 B CN 105636062B
- Authority
- CN
- China
- Prior art keywords
- node
- service
- information
- frequency range
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 230000005540 biological transmission Effects 0.000 title claims abstract description 69
- 238000000034 method Methods 0.000 title claims abstract description 50
- 230000019771 cognition Effects 0.000 title claims abstract description 24
- 239000011159 matrix material Substances 0.000 claims abstract description 71
- 230000009471 action Effects 0.000 claims abstract description 63
- 238000001228 spectrum Methods 0.000 claims abstract description 30
- 230000006978 adaptation Effects 0.000 claims abstract description 9
- 238000013461 design Methods 0.000 claims abstract description 7
- 230000008569 process Effects 0.000 claims description 20
- 239000013598 vector Substances 0.000 claims description 20
- 230000000694 effects Effects 0.000 claims description 16
- 238000009826 distribution Methods 0.000 claims description 7
- 239000003795 chemical substances by application Substances 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 6
- 230000002123 temporal effect Effects 0.000 claims description 6
- 230000006399 behavior Effects 0.000 claims description 5
- 101001115724 Homo sapiens MORN repeat-containing protein 2 Proteins 0.000 claims description 3
- 102100023291 MORN repeat-containing protein 2 Human genes 0.000 claims description 3
- 230000003542 behavioural effect Effects 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000008447 perception Effects 0.000 claims description 3
- 230000003068 static effect Effects 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 2
- 239000012634 fragment Substances 0.000 claims description 2
- 230000002441 reversible effect Effects 0.000 claims description 2
- 230000003595 spectral effect Effects 0.000 claims description 2
- 238000007726 management method Methods 0.000 abstract description 5
- 238000010276 construction Methods 0.000 abstract description 3
- 230000008859 change Effects 0.000 description 12
- 230000001149 cognitive effect Effects 0.000 description 7
- 230000006870 function Effects 0.000 description 6
- 238000004891 communication Methods 0.000 description 5
- 238000006243 chemical reaction Methods 0.000 description 3
- 230000004069 differentiation Effects 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 2
- 238000013475 authorization Methods 0.000 description 2
- 238000011217 control strategy Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- 241000208340 Araliaceae Species 0.000 description 1
- 241001269238 Data Species 0.000 description 1
- 235000005035 Panax pseudoginseng ssp. pseudoginseng Nutrition 0.000 description 1
- 235000003140 Panax quinquefolius Nutrition 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 230000002153 concerted effect Effects 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000005538 encapsulation Methods 0.000 description 1
- 235000008434 ginseng Nutrition 0.000 description 1
- 230000002452 interceptive effect Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000007480 spreading Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- 230000001550 time effect Effects 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W16/00—Network planning, e.g. coverage or traffic planning tools; Network deployment, e.g. resource partitioning or cells structures
- H04W16/14—Spectrum sharing arrangements between different networks
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W72/00—Local resource management
- H04W72/04—Wireless resource allocation
- H04W72/044—Wireless resource allocation based on the type of the allocated resource
- H04W72/0453—Resources in frequency domain, e.g. a carrier in FDMA
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B2201/00—Indexing scheme relating to details of transmission systems not covered by a single group of H04B3/00 - H04B13/00
- H04B2201/69—Orthogonal indexing scheme relating to spread spectrum techniques in general
- H04B2201/692—Cognitive radio
Abstract
The present invention relates to a kind of cognition wireless networks that service-oriented moderately services to transmit learning method, includes the following steps: step 1, acquires real time business and precedence information；Step 2, white space matrix and interference matrix are established；Step 3, according to node actionable course of action, business transmission time, routing and time delay are calculated；Step 4, the distributed collaboration design of local information and priority is carried out；Step 5, information exchange freedom degree and freedom degree restrictive condition are obtained；Step 6, by cooperative transmission and information exchange, multi-service application is carried out from adaptation；Step 7, learnt according to the Distributed Mandatory unrelated with model, establish frequency spectrum switching；Step 8, it is checked whether to meet user demand according to service quality.The present invention improves the performance of diversity real-time requirement business on multi-hop cognition wireless network using the Distributed Mandatory study unrelated with model, a kind of comprehensive business datum cross-layer management method of information exchange freedom degree construction is proposed, so as to have business pervasive for whole network.
Description
Technical field
The present invention provides a kind of data transmission method that teleservice moderately services, especially with regard to cognition wireless network
The middle cognitive process method using information exchange freedom degree belongs to cognition wireless network and designs applied technical field.
Background technique
Cognitive radio is the sensitive wireless telecom equipment of frequency-for having dynamic spectrum access, and great potential swashs
It has sent out and has sought better spectrum management and shared policy in community in engineering, economy, supervision, be under future wireless system field
One major issue.The cognition wireless network constructed based on cognitive radio, be by support cognitive radio technology terminal,
The wireless network of relevant rudimentary facility, control strategy composition.Cognition wireless network has a series of perceptions, in each mistake
Cheng Zhongke is with the space time information of collection network environment, including every layer from each Node extraction network of Parameters variation, really with this
Fixed most correct network parameter, meets individual node in network, end-to-end or even a group node communication target demand.
The effect for the wireless network for supporting terminal, relevant rudimentary facility and the control strategy of cognitive radio technology to form
It is affirmed fully in multiple fields, such as smart grid, device-to-device network, military communication, GPS positioning, medical body
Educate net etc..The type of business and QoS requirement of future mobile communications are more and more diversified, and the service ability of network is
It is limited, which results in the gap between business demand and network capability is increasing, it is final will be so that network be difficult to prop up
Hold diversified business.The research focus of cognitive radio is concentrated mainly on dynamic spectrum access, but cognition wireless network is answered
It was not hashed out with the performance of layer.Some important problems, such as what application can make full use of frequency spectrum white space,
And whether existing wireless protocols are capable of providing satisfied performance, are not resolved still.The business of " bandwidth-starvation " is answered
With being the basis for giving full play to cognitive radio potentiality, therefore, promote to how to improve end-to-end efficiency in cognition wireless network
Discussion into service application problem is extremely important.
However, there are two technological challenges: first challenge in the wireless context in the above problem, it is full in the horizontal direction
Sufficient diversity business demand, this needs spatial and temporal distributions, the priority of the different business amount for frequency spectrum resource, then makes full use of
Channel, interference, business dynamic change characterization, diversity business demand is analyzed, finally identifies available frequency spectrum cavity-pocket.
Second challenge, be in vertical direction realize each layer of network architecture (physical layer, MAC layer, routing layer, link layer) across
Layer optimization, it is true that this needs the demand of a suitable frequency allocation and speed self-adaption method to secondary user (unauthorized) to carry out
Recognize, in the case where not interfered by primary user (authorization), meets time user's QoS requirement.One side of novel radio real time business
Face needs to safeguard a large amount of shared datas and control information, these data, information are often of short duration；On the other hand its activity has very
Strong instant binding feature.For features described above, need according to spectrum opportunities, disturbed condition and information interests, to wireless real-time
Cross-layer in business distributes Method of Spreading Design, is allowed to have autonomous suitable change characteristic.
Summary of the invention
In order to establish dynamical service mechanism, the present invention both horizontally and vertically above passes through justice in proportion above-mentioned,
Come capturing service data and information exchange freedom degree, adaptively to match resource, and with whole net availability of frequency spectrum highest
The Land use systems of optimal multidimensional service type information are sought for criterion.Target is the energy construction one between network and business
Kind comprehensive business datum cross-layer management method, so as to have business pervasive for whole network.
The invention proposes the cognition wireless networks of a real-time diversified business to transmit learning method, and Lai Shiying business is dynamic
Inner link between state variation characteristic establishes the wireless transmission of the end-to-end efficiency of multi-criteria optimization.The present invention can answer extensively
For in environment of cognitive radio network.
To achieve the goals above, the present invention adopts the following technical scheme that:
(1) real time business and precedence information are acquired；
(2) white space matrix and interference matrix are established；
(3) according to node actionable course of action, business transmission time, routing and time delay are calculated；
(4) the distributed collaboration design of local information and priority is carried out；
(5) information exchange freedom degree and freedom degree restrictive condition are obtained；
(6) by cooperative transmission and information exchange, multi-service application is carried out from adaptation；
(7) basis and model it is unrelated Distributed Mandatory study, establish frequency spectrum switching；
(8) it is checked whether to meet user demand according to service quality.
In the step (1), under multi-hop environment, cognition wireless network is made of primary user, secondary user, network trunk.Recognize
Know that wireless network includes 3 set, i.e. primary user's collection, a network node including time user and network trunk transmission
The network link collection of collection and a connection time user and network trunk transmission.There are multiple nodes and multilink, node in network
It is time user or is relay transmission node.In order to avoid primary user's interference, network node is passed under frequency spectrum cavity-pocket
Defeated, therefore, network node will establish a neighbor node link, and node is only with available frequency band collection.
The otherness of service feature needs to be conceived to the object (user) of any wireless network services and pair of wireless network transmitting
As (information).Demand of the business to network is more and more various and changeable, and the service resources of network node are described with traffic matrix, packet
Include the routing of different links and frequency range composition.In traffic matrix, 1 indicates that designated frequency band is available and link and node is connected, 0 table
Show that frequency range is unavailable or connectionless.Service resources matrix depends not only on network topology connection, and depends on identical frequency range
On Communication Jamming.
User's real time business application of the network carrying have it is multiple, and by the data packet group of multiple priority at wherein highest
Priority is directed to the real time business of primary user.Correspondingly, remaining priority can use a triple, i.e. average data packet is long
Degree, influence power, Packet Delay off period determine.Priority determines frequency spectrum access capability.Most in addition to primary user's frequency range
High priority waits for an opportunity the priority for accessing frequency range there are also time user.Primary user has highest priority, therefore connects at any time
Enter frequency range；On the contrary, secondary user is only capable of access frequency spectrum cavity-pocket.Due to being influenced by multihop network environment, such as interference, signal are strong
Degree, multi-hop, etc., secondary user, which can perceive primary user, to weaken, and secondary user also can generate interference to primary user, these interference rely on
In the position of primary user.
In the step (2), it is assumed that spectrum opportunities are available to secondary user, and primary user can prevent all neighbours' links using frequency
Section, the node perceived channel, and obtain the white space matrix of primary user.In white space matrix, 1, which indicates that primary user occupies, is specified
Frequency range, and link and primary user's interference, are otherwise 0.
In the interference matrix of priority, 1 indicates to be interfered by priority level if link is using designated frequency band, otherwise for
0.Under wireless environment, neighbours' link is interfering with each other, and network node can be interfered with each other with other relay transmission nodes.In order to carry out
Coordinate, interference matrix is calculated by the information exchange of neighbor node.In addition, the white space matrix of higher priority and interference square
Battle array will affect available service matrix.
It needs to consider following two situation when establishing interference matrix.In the case of the first, network node can transmit simultaneously,
Receive differentiated service, node cannot reuse the frequency range that neighbor node uses.If frequency range is used by neighbor node, frequency range
In all elements of interference matrix be set to 1.Under second situation, network node cannot transmit simultaneously, receive differentiated service.
If all elements in interference matrix in frequency range are set to 1 using designated frequency band.Moreover, if neighbor node uses
Network link, then regardless of which kind of frequency range used, all elements in node interference matrix are also set to 1.
In the step (3), in order to realize that real time business relays, network node has an actionable course of action.Network trunk
The link set of transmission selection neighbor node indicates that network node is selected with one user service vector to take concerted action
The probability of above-mentioned action.In this case, the link and frequency range of selection will not be interfered with the communication of higher priority.
In order to calculate the real-time service transmission time under priority, under designated frequency band, need to combine node on the link
Transmission rate and Packet Error Ratio, and estimated by the adjustment of the link of physical layer and MAC layer.Each " link-frequency range " pair
Channel condition is indicated that primary user is according to the synchronous knot of a timeslice by the continuous time Markov Chain of a finite state number
Structure utilizes frequency range, and discrete markoff process is obeyed in the use of frequency range.
Motion vectors are the vector of the all-network relay transmission node of real time business.Real time business sends number from source node
Destination node is arrived according to packet, " link-frequency range " that routing i.e. data packet passes through is to set.
If the action of relay transmission node changes, the routing of real time business can also change.According to above topology, often
A relay transmission node for receiving data packet can determine wherein relay transmission data packet, and which kind of frequency range reduce end using
To terminal delay time.Then, source node needs to obtain the Delay of other nodes according to the action of relay transmission node.
In the step (4), real time business aims at minimum end-to-end time delay.Distributed method and centralization side
Method is compared, can be by periodically collecting local information, therefore preferably meets multi radio, multiband equipment environment.More marks
Quasi- optimization solution depends on the time delay of different data packet multi-hop transmissions, and the data transmission of source node needs to be relayed in time.
In order not to collect interconnection information in source node, using a distributed multi-criteria optimization scheme come the sheet of collector node
Ground information, (the i.e. information of multi-hop environment interior joint itself, including various businesses parameter and time parameter.In wireless network, need
Local information is first collected, then carries out the business data transmission of multi-hop.) make the time delay expectation of differentiation real time business application most
It is small.In each node, the time delay end to end for sending data packet needs to decompose.For calculation delay expectation, network node needs
There is an estimation propagation delay time.Meanwhile node needs to keep and update the time delay vector of each priority, wherein first is preferential
Grade is that primary user retains.
In the step (5), local information is swapped using a Cooperation controlling frequency range.The data of secondary user service
Transmit temporally fragment, the cooperation interval including timeslice duration and network node.Each node is in timeslice and assists
Action is selected after being spaced.Cooperation interval is in addition to comprising white space matrix and interference matrix, also including time delay vector sum RTS/CTS
The control message of cooperation.Cooperation interval in each timeslice can wait for an opportunity access to frequency range and provide actionable course of action collection and data packet biography
Defeated relay transmission selection.Different local informations has different influences to objective function.
Next, needing to obtain the return interests of information in associated ratings.In static network, if node receives letter
Breath, then information report, which is deemed to be, it is expected brought interests according to time delay.By calculating available optimal time delay expectation,
It is the decreasing function of a hop count, i.e., time delay period hopes smaller that information is collected more.Reducing expectation time delay can be returned by information
Report is quantified.When more nodal informations reach decision node, information is likely to out-of-date.Due to the spatial and temporal distributions of portfolio
The dynamic characteristic showed, thus information cannot in dynamic environment accurately reaction network position.Once information is out-of-date, collect
Out-of-date information is not used, that is, there is no interests.Therefore, it is necessary to use ginseng of the freedom of information degree as searching with maximum scores
Amount.Assuming that freedom of information degree is only a function of network change speed, information exchange needs preferential by the highest in secondary user
Grade determines.Freedom of information degree is the maximum hop count of information transmission, and the network is deemed to be constant in this way, and any network change
It is ignored in interval.
Assuming that the network node in freedom of information degree constitutes an information unit, neighbor node is can be mutual with network node
The mutually node of interference, meeting and freedom degree range in certain jump are inconsistent.If all neighbor node centainly jump from
In spending, then all information can be transmitted in node in time；Otherwise, too far neighbor node cannot pass in time
Send interference information into node, such case, that is, information exchange mismatches.
Cost brought by information exchange, i.e., increased expectation time delay.In timeslice, there is an information constantly to exchange
The number of decision is carried out in interval and a learning process.Information exchange is carried out in specified cooperation frequency range, has simple day
The network node simultaneous transmission data and control signal of line, are impossible.Therefore, the time cost of information exchange can subtract
Few node uses the effective transmission rate of link and frequency range.Increasing effective transmission time can be such that the performance of diversity requirement business occurs
Degrade, which depends on the content and network change speed of local information exchange.
In the step (6), under multi-hop environment, distributed resource of the invention is assisted from adaptation method by information exchange
Diversity real-time requirement business between tune adjacent node.Firstly, node is by the influence power in priority triple from service coverage
Data packet is selected, which is surrounded by an action.Meanwhile time delay off period in priority triple, data packet length
Application layer message is transmitted in the network layer of this action.Then, network condition is transmitted in physical layer and MAC layer, and is calculated
Business transmission time.
In addition to transmission rate and Packet Error Ratio, the interference that neighbor node action generates will affect action selection, so neighbours save
The information that point receives is in freedom of information degree.Node obtains estimation time delay and the neighbours that destination node is reached from neighbor node
The action of node and interference matrix.According to the Delay of neighbor node, network node can update it and reach different purpose sections
The estimation time delay of point, and can identify the action of minimal time delay.
Then the adaptive with being divided into cooperative transmission process and information exchanging process two parts of multi-service application.Wherein, it passes
Defeated cooperating process selects optimal action: initial phase according to 3 stages, it is thus necessary to determine that data to be transferred packet；It reports class, needs
Actionable course of action collection is established, and estimates frequency range condition, then calculation delay it is expected；Negotiation phase was needed according to time delay by the phase,
Select optimal strategy.Information exchanging process is then according to initial phase, study stage, traffic matrix estimation stages, interference matrix
This 4 stages of establishment stage collect local information.Actionable course of action collection depends on the action of other neighbor nodes, and network node needs
The behavior of network node is carried out from adaptation using learning method, and reduce the complexity of dynamic adjustment.
In the step (7), in order to improve real-time service performance and adapt to the dynamic characteristics of portfolio spatial and temporal distributions, cognition
Wireless network needs multiple autonomous nodes, i.e. intelligent body (Agent), each node can be adjusted according to local information, do
Transmission decisions and negotiation of radio resources.The prior art ignores the joint behavioral characteristics between node, still, neighbor node action
Description, study and activity but play key in the high efficiency of diversity requirement business and robustness with the variation of time
Effect.Therefore, the present invention is according to knowledge communication network dynamic change, and adjustment cross-layer transmission decision efficiently to utilize share,
The Internet resources of time-varying.For the differentiation appropriateness service environment on wireless network, which can improve the efficiency of main body.
A smart frequency spectrum switchover policy is designed, wherein adjusting time user's collection to provide end-to-end appropriateness service.This can be with
It is realized by the way that frequency spectrum handover decisions are used one as the markov decision process of activity return, and needing to learn can every trade
Dynamic collection.Actionable course of action collection is determined according to information exchange, and for the behavior, can learn row of the neighbor node in information unit
For.Then, learnt using Distributed Mandatory, to find optimizing decision policy.This method does not need conversion and reward model, so that
Secondary user finds optimal policy.I.e. using one for action sequence, and connects and live according to secondary user's connection status, secondary user
Dynamic, secondary user connects the on-line study of return, secondary user connection, resource matched to meet high energy efficiency.
Virtual action utilizes domain knowledge comprising rewards and punishments information and actionable course of action collection by one.Actionable course of action collection foundation
Interference matrix and white space matrix.Main body selects suitably to act the spatial and temporal distributions performance for going to portfolio in each state,
Such as time delay, priority in proportion in fair, packet loss, make it be converted to a new state.On the one hand, main body is according to having wanted
At task judged, obtain range adjustment；On the other hand, main body is according to the priori knowledge of action effect, when carrying out minimum
Prolong action selection.In order to which the exchange of Business Stream and control stream is better achieved, need one based on strategic vector determination and interferes
The adjustment that matrix determines.
In the step (8), is checked whether to meet user demand according to service quality, need pair from any wireless network services
The two aspect expansion of object (information) transmitted as (user) and wireless network is examined.If being able to achieve service adaptation, terminate to flow
Journey；Otherwise, restart to execute from acquisition information exchange freedom degree step.
The present invention compared with prior art, has the following beneficial effects:
1, the invention proposes the data transmission schemes that a differentiated service moderately services, and using unrelated with model
Distributed Mandatory learns to improve the performance that real time business is applied on multi-hop cognition wireless network.
2, the present invention devises information exchange to learn spectrum opportunities, multiple services dynamic for autonomous secondary user collection
Under state spatial and temporal distributions, method proposed by the present invention can meet the QoS requirement of general multi radio network well.
3, the present invention defines information exchange in adaptive virtual action to reduce the desired influence of end-to-end time delay
Freedom degree, a kind of comprehensive business datum cross-layer management method of construction, so as to have business pervasive for whole network.
Detailed description of the invention
Fig. 1: for overall structure flow diagram of the invention
Fig. 2: for the priority scheduling figure of secondary user's differentiated service collection of the invention
Fig. 3: for the transmission time figure of timeslice and node of the invention
Fig. 4: for freedom of information degree autonomous learning flow chart of the invention
Specific embodiment
In order to keep the above objects, features and advantages of the present invention more obvious and easy to understand, with reference to the accompanying drawing to this hair
Bright specific embodiment is described in detail.
Present invention will be further explained below with reference to the attached drawings and examples.
The cognition wireless network that a kind of service-oriented provided by the invention moderately services transmits learning method, including walks as follows
It is rapid:
(1) real time business and precedence information are acquired；
In order to embody customer-centric, cognition wireless network is described with G (P, N, E), and primary user collects P={ P1,…,PQ,
Network node collection N={ n1,…,nN, network link integrates as E={ e1,…,eL}.There are N number of node and L link in network, this
A little nodes or one user or a relay transmission node.Frequency range in network integrates as M={ M1,…,MQ, Q is frequency
Section sum.The business datum of actual environment is obtained for the method for using multi-point sampling, distribution multiple, network node n's
Traffic matrix Sn=[Sij]∈{0,1}L×QIt is expressed as follows:
There are larger differences for demand of the different classes of user to business.In order to realize that high-precision type of service feature mentions
It takes, enables RiFor the real-time traffic demands of secondary user i, RiBy GiThe data packet group of a priority is at shared U business.Setting is shared
G priority, priority set C={ C1,…,CG, wherein primary user's (authorization) occupies highest priority C1.Accordingly, secondary use
Family (unauthorized) grade Cg(g > 1) uses tuple (Lg,ρg,Dg) be described and determine: where LgFor grade CgIn average
According to packet length, ρgFor grade CgInfluence power, DgFor grade CgThe time delay off period of middle data packet.Influence power can distinguish real-time industry
The order of priority for demand levels of being engaged in, if g < g ' (g=2 ..., G), has ρg≥ρg′.Finally, influence power is encapsulated, and is placed on
Data packet header.
As shown in Fig. 2, multimedia transmission (such as wireless device to equipment application, distributed search, distributed flow media) is
The typical example that business moderately services.The information transmitted needed for wireless network can be carved by two dimensions of user and business
It draws, user first generates diversified information requirement, then re-maps as diversified business.Real-time traffic demands R1,R2,RUIn conjunction with
Frequency range M1,…,MQ, collect P according to primary user1,…,PQ, the virtual team of different frequency range collection is generated by service coverage and time user service
Column.Secondary user service is realized after experience frequency spectrum switching, needs to be sent according to the influence power for being encapsulated in packet stem most important
Data packet.
(2) white space matrix and interference matrix information are established；
Network node n passes through frequency spectrum perception, the white space matrix O of acquisitionn=[Oij]∈{0,1}L×QIt is expressed as follows:
Priority CgThe interference matrix of (g >=2) is In=[Iij]∈{0,1}L×Q:
Correspondingly, traffic matrixIndicate network node n in grade CgThe differentiation industry of middle received higher priority interference
Business resource, then Representing matrix reverse turn operation.
(3) according to node actionable course of action, business transmission time, routing and time delay are calculated；
The action of network node n is An=(e ∈ En,M∈Mn).Assuming that a network trunk n can select the chain of neighbor node
Road collectionThe secondary user service vector of network node n is sn=[sA| A=(e ∈ En,M∈Mn)], network node n selection
The probability of action A is sA.A=(e, M) isAvailable resources, i.e.,Element SeM=1, grade CkInterior joint n's is all
Actionable course of action collection isIt can be expressed as
In view of the information content has stringent timeliness, each network node n calculates the real-time service transmission timeIn priority CkUnder, e ∈ En,M∈Mn:
Transmission rate and Packet Error Ratio of the frequency range M on link e are used with node n is respectively indicated.Assuming that each " chain
The channel condition on road-frequency range " pair can be described by the continuous time Markov Chain of a finite state number.Q frequency
Discrete markoff process is obeyed in the use of section.The network state of timeslice t isIndicate the state of frequency range q idle (0) or busy (1).
Motion vectors Ai=[An|n∈θi] it is RiAll-network relay transmission node vector.Assuming that RiFrom source nodeSend qiA data packet is to destination nodeSo its routing is defined as ψi={ ψij| j=1 ... qi, wherein
ψijFor RiIn j-th of data packet routing.Route ψijFor link-frequency range pair set that data packet passes through, i.e.,
θij=(e, M) | RiIn j-th of data packet using frequency range M pass through link e
If the action of a relay transmission node changes, corresponding relay transmission RiRouting ψij(Ai) can also change,
Using routing ψij(Ai) data packet transmission delay be dij(ψij(Ai)).According to the topological structure, the net of data packet is each received
Network relay transmission node can determine wherein to relay data packet, and which kind of frequency range reduce time delay end to end using.In order to
Data packet transmission delay is calculated, source node needs to obtain the Delay of other nodes according to the action of relay transmission node, i.e.,Ri∈Cg。
(4) the distributed collaboration design of local information and priority is carried out；
In each node n, a data packet j ∈ C is sentgEnd-to-end time delay can be carried out following decomposition:
Indicate data packet j ∈ CgThe time delay expectation of destination node is reached from node n,For data
It wraps j and reaches the time delay undergone before node n.Influence power ρgService coverage determine send data packet j ∈ Cg。ρgInformation encapsulation
To data packet header, andIt can be calculated according to the timestamp in data packet header.The priority scheduling device of each node
Ensure that high priority is not influenced by low priority.Since in node nIt is fixed, then cross-layer multi-criteria optimization problem becomes
At:
Wherein,Indicate grade CkIn data packet from relaying transmission node n reach destination node
Time delay expectation.ξ, which indicates one, makes probabilityThe personal distance of very little.For
The expectation of assessment time delayNeed an estimation propagation delay timeSet each node n
Keep and update the time delay vector of each priorityWherein first it is excellent
First grade is primary user's reservation.
(5) information exchange freedom degree and information unit restrictive condition (information interests, freedom of information degree) are obtained；
As shown in Figure 3, it is assumed that local information LnIt is swapped using a Cooperation controlling frequency range.Transmission temporally divides
Piece enables tsFor timeslice duration, dc(Ln) be network node n cooperation interval.Each node is in timeslice between cooperation
Every rear selection action An.In addition to white space matrix and interference matrix, cooperation interval also includes time delay vector dnWith control message RTS/
CTS.Cooperation interval in each timeslice can wait for an opportunity access to frequency range and provide actionable course of action collectionIn data packet transmission
It is selected after transmission.
Different local information LnFor the objective function of reductionThere is different influences.It enablesFor neighbor node collect local information, it apart from node n have x jump away from
From,Indicate the node collection for having x hop distance with node n.Enable Ln(x)={ In(l) | l=1 ..., x } it is all neighbor nodes
The local information being collected into.For local information, optimal time delay is enabled to be desired for
X is bigger, and time delay it is expected Gn(g, x) is smaller.
Enable Vn(g,InIt (x)) is grade CkMiddle information In(x) return interests.In static network, Vn(g,In(x)) it defines
Are as follows: Vn(g,In(x))=Gn(g,x-1)-Gn(g,x)。
Because of Ln(1)=In(1), so Vn(g,In(1))=Gn(g,1).If node n receives information In(x), then believing
Breath return Vn(g,In(1)) it is deemed to be and it is expected according to time delayBrought interests.For information In(x), optimal
Time delay it is expected Gn(g, x) may be expressed as:
Once information is out-of-date,Collecting out-of-date information there is no interests.In Dynamic Networks
In network, onceSo for x≤x '≤hnHaveTherefore, freedom of information degree is defined
DoF (k, v) are as follows:
DoFn(k, v) ≡ argmaxx,
Wherein λ (g, v) >=0 indicates the variation of real time business bring minimal time delay, which determines in grade CkIn
Receive the minimum interests of local information.DoFn(g, v) depends on the mobile pace of change v of wireless network.In a mobile environment, it needs
There is a higher threshold value φ (g, v), so that information In(x) still valuable and be used to exchange.Then smaller letter is generated
Cease freedom degree DoFn(g, v), wherein freedom of information degree DoFn(g, v) changes with different priority.It is compared with low priority,
High priority has more Internet resources, if g < g ', threshold value λ (g, v)≤λ (g ', v), then there is DoFn(g,v)
> DoFn(g′,v).That is a high priority CgFreedom of information degree DoFnOne low priority C of (g, v) ratiog′Freedom of information degree
zn(g ', v) is big.
Assuming that freedom of information degree is only a function of network change speed v, i.e. DoFn(g, v)=DoF (v).Information is certainly
It needs to be determined by the highest priority in secondary user by degree DoF (v).Freedom of information degree DoF (v) is defined as the information in τ and passes
Defeated maximum hop count, the in this way network are deemed to be constant, and any network change in η (the v)≤1/v of interval can be neglected
Slightly.
Referring now to for cost brought by information exchange, that is, it is expected the increase of time delay.In timeslice, ts(v) it is
The interval that information constantly exchanges in network.Defining in η seconds has c timeslice, i.e. ts(v)=η (v)/c, c be learning process in into
The number of row decision.It is compared with η, each tsIt is short enough with the duration of timeslice, therefore negligible tsIn network change.
(6) by cooperative transmission process and information exchanging process, multi-service application is carried out from adaptation；
Node n is by influence power ρ firstgA data packet j ∈ C is selected from service coverageg, which is surrounded by an action
An.Then Cg,Lg,DgApplication layer message be transmitted in the network layer of this action.Then it is transmitted in physical layer and MAC layer
Network conditionAnd calculate business transmission time.
In addition to transmission rateAnd Packet Error RatioThe interference that these neighbor node action generate will affect action selection,
So the information that neighbor node receives is in freedom of information degree, wherein Ln(z)={ In(l) | l=1 ..., z }.- n (h) is net
Neighbor node collection of the network node n in the information unit that z is jumped, the local exchange information L of network noden(z)={ Ig(-n(z),
A-n(z)),A-n(z),d-n(z)}.Node n obtains the estimation time delay d that destination node is reached from neighbor node-n(z)And neighbor node
Action A-n(z)With interference matrix Ik(-n(z),A-n(z)).According to the Delay d of neighbor node-n(z), network node can be more
The estimation time delay of different destination nodes is newly reached, and can identify the action of minimal time delay.Then, periodical in each timeslice
Ground repeats following two process:
1. in cooperative transmission, in initial phase, it is assumed that data packet j ∈ Cg, Cg,Lg,These business information
It is calculated in application layer, according to influence power, distributes a data packet j buffered on node n to be transmitted；Reporting class,
It needs for priority C in handshake interfacegTraffic matrix on node nEstablish actionable course of action collectionThen
It is adjusted by link, provides link frequency range to upper transmission rate to physical layer and MAC layerAnd Packet Error RatioNegotiating rank
Section, ifThe data packet is then lost, minimal time delay routing and Frequency Band Selection is otherwise found, that is, exists
Actionable course of action collectionIt is upper to find optimal action
In the report stage, grade CkEach actionTime delay expectation can be calculated:Wherein,For neighbor node n ' (An)
In grade CkIn time delay vector d-nOn element.And in negotiation phase, after determining next relaying and frequency range, sends and determine row
Dynamic informationRTS request into next relay transmission, wait CTS react and transmit data packet.After selecting optimal action,
Update estimation time delayThen the time delay vector updated is provided to handshake interface.
2. network node n collects white space matrix and neighbor node in information list in the information exchange, in initial phase
Information I in memberg(-n(z),A-n(z)),A-n(z),d-n(z)；It can be adjacent to node n by neighbor node action in the study stage
The behavior for occupying node is confirmed, or uses strategic vector s (n ')=[sA(n ') | A=(e ∈ En′,M∈Mn′)] obtain more
Good secondary user service, wherein n ' ∈-n (z), the s in strategic vectorA(n ') indicates the probability of node n ' selection one action A；
In traffic matrix estimation stages, in the white space matrix and interference matrix I of neighbor node n ' collectionk(n′,An′) in, it is each preferential
Grade obtains traffic matrixAnd available resources are provided to network layer real-time resource management moduleIn interference matrix establishment stage, more new information will be according to the action of newest selection, at the latest time delay variable and interference square
Battle array Ig(n,An).Finally, interference matrix and the node in Predefined information freedom degree swap.
(7) basis and model it is unrelated Distributed Mandatory study, establish frequency spectrum switching.
Next one smart frequency spectrum switchover policy of design is adjusted by secondary user to provide end-to-end appropriateness and service.Frequently
Spectrum handover decisions can use a markov decision process as activity return to realize.According to information exchange Ln(z), and
Actionable course of action collection is determined for the behaviorIt can learn behavior of the neighbor node in information unit.
The markov decision process of one finite state is expressed as a tuple (S, A, T, R), and wherein S is system mode
Collection；A is the candidate action collection of each state；T={ Ps,s′It (a) } is condition transition probability collection, wherein Ps,s′(a) for when in state s
The middle state transition probability using action a and from state s to s '；AndIt indicates when using activity a ∈ A and s
Expense when ∈ S.The markov decision process is made of following iterative step: firstly, intelligent body (Agent) perceives environment
And observe s ∈ S；Secondly, being based on s, intelligent Agent selection activity a ∈ A is executed in the environment；Again, environment is from new state s '
It is converted, and generates return (expense) r ∈ R；Then, intelligent Agent, which receives, returns and is used to more new departure；Finally,
It repeats the above process.
If it is known that the conversion and return of markov decision process, it will be able to obtain each time using activity iteration and use
The optimum state at family.Its complexity is approximately n2, n is status number.The above method is very effective when state space is small, still
As incremental and behavioral characteristics, the probability transfer function of complexity are difficult to derive.It is therefore desirable to using one and model without
The Distributed Mandatory learning method of pass, to find optimizing decision policy.This method is not needed to convert and be returned, so that secondary user seeks
Look for optimal policy MOPT (s) ∈ A, i.e., the one action sequence { a for s1,a2,a3..., to maximize all expectation foldings for a long time
Button return (or reducing expense):
1. secondary user's connection status: the secondary user connection given for one, the network state table before i+1 time is interrupted
It is shown asMkIt is interrupted for i+1 time when occurring, frequency range when secondary user service.Indicate frequency range Mk
The condition of Packet Error Ratio.WithThe secondary user of type priority grade j experience i-th interruption is respectively indicated in frequency range MkOn arrive
Up to rate and service time.
2. secondary user's connection activity: when secondary user's disconnecting, needing to select to stay in current frequency range, or exchange to another
In one available frequency band.It enablesIt is secondary user in state sj,iAnd candidate active when i+1 time interruption.For Frequency Band Selection parameter, it is used to determine selection frequency range MkAs transmission band and the probability when i+1 time is interrupted.It is current
When frequency spectrum handover delay being hoped to be more than the time delay off period, data packet can lose.
3. secondary user connects return: for switching, the prediction that a movable return r is defined as multimedia transmission is end-to-end
Efficiency.It is expected that end-to-end efficiency consists of two parts, i.e., due to the Packet Error Ratio of channel condition and when frequency spectrum switching expectation time delay is super
Cross the packet loss of time delay off period.By this Reward Program, mandatory learning is used to maximize end-to-end efficiency, and balancing spectral
Time delay caused by switching.The expectation time delay of applied business high priority, will not be by the secondary customer impact of other low priorities.But
Be if secondary user belongs to low priority, can consider using the influence of high-priority traffic as end-to-end performance criteria, and according to
Queuing model priority-based.
4. the on-line study of secondary user's connection: target of the intelligent Agent when i+1 time is interrupted is that searching one is optimal
Activity, in current policies MOPT (sj,i,aj,i) under maximize end-to-end efficiency.
(8) it is checked whether to meet user demand according to service quality.
It is adjusted for the diversification of type of business and QoS requirement according to above-mentioned service adaptation situation, or
Terminate process, or carries out the adjustment of information exchange freedom degree DoF (k, v).
The part not illustrated in specification is the prior art or common knowledge.The present embodiment is merely to illustrate the invention,
Rather than limit the scope of the invention, the modifications such as equivalent replacement that those skilled in the art make the present invention are considered
It falls into invention claims institute protection scope.
Claims (7)
1. a kind of cognition wireless network that service-oriented moderately services transmits learning method, it is characterised in that: this method include with
Lower step:
Step 1: establishing service resources matrix for each network node, the excellent of real time business is defined according to the otherness of real time business
First grade；
The service resources matrix includes the routing being made of different links and frequency range；In matrix 1 indicate designated frequency band it is available and
Link and node is connected, and 0 indicates that frequency range is unavailable or connectionless；The traffic matrix S of network node nn=[Sij]∈{0,1}L×QTable
Show as follows:
Wherein L is link sum, and Q is frequency range sum, link eiBelong to link set E={ e1,…,eL, frequency range MjBelong to frequency range collection M
={ M1,…,MQ}；
Step 2: establishing white space matrix according to whether primary user occupies a certain frequency range and whether interference occurs with a certain link；
Whether high priority link is interfered to establish interference matrix using a certain frequency range according to link；The white space matrix and interference matrix
The behavioral characteristics that spatial and temporal distributions for describing portfolio are presented；
Network node n passes through frequency spectrum perception, the white space matrix O of acquisitionn=[Oij]∈{0,1}L×QIt is expressed as follows:
Priority CgThe interference matrix of (g >=2) is In=[Iij]∈{0,1}L×Q:
Correspondingly, traffic matrixIndicate network node n in grade CgThe differentiated service money of middle received higher priority interference
Source, then Representing matrix reverse turn operation
Step 3: according to node actionable course of action, business transmission time, calculating real time business from source node and transmit packets to target
The routing of node and estimation time delay dij(ψij(Ai)) it is expressed as follows:
Wherein, ψijFor RiIn j-th of data packet routing, Ai=[An|n∈θi] it is RiAll-network relay transmission node row
Moving vector, STngIt is network node n in grade CgIn take action vector AiBusiness transmission time；
Step 4: the distributed collaboration design of local information and priority is carried out, cross-layer multi-criteria optimization problem is become:
Wherein,Indicate grade CkIn data packet from relaying transmission node n reach destination node when
Delay to hope, ξ, which indicates one, makes probabilityThe personal distance of very little, DgFor etc.
Grade CgThe time delay off period of middle data packet,Time delay after undergoing routing before reaching node n for data packet j；
Step 5: obtaining information exchange freedom degree and freedom degree restrictive condition；
Temporally fragment is transmitted, t is enabledsFor timeslice duration, dc(Ln) be network node n cooperation interval, each node exists
Action A is selected in timeslice and behind cooperation intervaln；For local information, optimal time delay is enabled to be desired forX is bigger, and time delay it is expected Gn(g, x) is smaller；Enable Vn(g,InIt (x)) is grade CkIn
Information In(x) return interests；In static network, Vn(g,In(x)) is defined as: Vn(g,In(x))=Gn(g,x-1)-Gn(g,
x)；If node n receives information In(x), then information report Vn(g,In(1)) it is deemed to be and it is expected according to time delayBrought interests；For information In(x), optimal time delay it is expected Gn(g, x) may be expressed as:In dynamic network, onceSo for x≤x '≤
hnHaveTherefore, freedom of information degree DoF (k, v) is defined are as follows:
DoFn(k, v) ≡ arg max x,
Wherein λ (g, v) >=0 indicates the variation of real time business bring minimal time delay, which determines in grade CkIn receive this
The minimum interests of ground information；
Step 6: by cooperative transmission and information exchange, multi-service application being carried out from adaptation；
Two processes of cooperative transmission and information exchange of execution are repeated periodically in each timeslice, in actionable course of action collection
It is upper to find optimal actionUpdate estimation time delayThen information exchange is given
Interface provides the time delay vector updated；According to the action of newest selection, at the latest time delay variable and interference matrix Ig(n,An) update letter
Breath, and then interference matrix and the node in Predefined information freedom degree swap；
Step 7: being learnt according to the Distributed Mandatory unrelated with model, establish frequency spectrum switching；
Realize that smart frequency spectrum switches using a markov decision process as activity return, using one and model without
The Distributed Mandatory learning method of pass finds optimizing decision；
Step 8: checking whether that meet user demand transmits from the object and wireless network of any wireless network services according to service quality
Object two in terms of be unfolded examine；If being able to achieve service adaptation, terminate process；Otherwise, from acquisition information exchange freedom degree
Step restarts to execute.
2. the cognition wireless network that a kind of service-oriented according to claim 1 moderately services transmits learning method, special
Sign is: in order to realize high-precision type of service feature extraction, enabling RiFor the real-time traffic demands of secondary user i, RiBy GiIt is a excellent
The data packet group of first grade is at shared U business；It sets and shares G priority, priority set C={ C1,…,CG, wherein leading
User occupies highest priority C1；Accordingly, secondary user gradation Cg(g > 1) uses tuple (Lg,ρg,Dg) be described and determine: where
LgFor grade CgIn average data packet length, ρgFor grade CgInfluence power, DgFor grade CgThe time delay off period of middle data packet；
Influence power can distinguish the order of priority of real-time traffic demands grade, if g < g ' (g=2 ..., G), has ρg≥ρg′。
3. the cognition wireless network that a kind of service-oriented according to claim 1 moderately services transmits learning method, special
Sign is: need to consider following two situation when establishing interference matrix in step 2: (1), network node can be transmitted simultaneously, be connect
By differentiated service, node cannot reuse the frequency range that neighbor node uses, if frequency range is used by neighbor node, in frequency range
The all elements of interference matrix be set to 1；(2), network node cannot transmit simultaneously, receive differentiated service, if using referring to
Determine frequency range, then all elements in the interference matrix in frequency range are set to 1, moreover, if neighbor node uses network link, that
Regardless of which kind of frequency range used, all elements in node interference matrix are also set to 1.
4. the cognition wireless network that a kind of service-oriented according to claim 1 moderately services transmits learning method, special
Sign is: for time interval described in step 5 in addition to white space matrix and interference matrix, further including includes time delay vector dnAnd control
Message RTS/CTS, each timeslice tsIn cooperation interval, for frequency range wait for an opportunity access actionable course of action collection is providedIt is passed with data packet
Defeated relay transmission selection.
5. the cognition wireless network that a kind of service-oriented according to claim 1 moderately services transmits learning method, special
Sign is: cooperative transmission process described in step 6 selects optimal action by 3 steps:
Step 601a: initial phase determines data to be transferred packet；
Step 602a: report class establishes actionable course of action collection, and estimate frequency range condition, and then calculation delay it is expected；
Step 603a: negotiation phase selects optimal strategy according to time delay by the phase；
The information exchange passes through 4 collection step local informations:
Step 601b: initial phase, network node collect the information of white space matrix and neighbor node in information unit；
Step 602b: the study stage confirms the behavior of nodes neighbors node by neighbor node action, or uses
Strategic vector obtains preferably time user service；
Step 603b: traffic matrix estimation stages, it is each preferential in the white space matrix and interference matrix that neighbor node is collected
Grade obtains traffic matrix, and provides available resources to network layer real-time resource management module；
Step 604b: interference matrix establishment stage, more new information will be according to the action of newest selection, at the latest time delay variable and interference
Matrix；Finally, interference matrix and the node in Predefined information freedom degree swap.
6. the cognition wireless network that a kind of service-oriented according to claim 1 moderately services transmits learning method, special
Sign is: the switching of frequency spectrum described in step 7 is realized by a markov decision process as activity return, and is used
One Distributed Mandatory learning method unrelated with model finds the optimum state of time user, according to secondary user's connection status,
Secondary user's connection activity, secondary user connect the on-line study of return, secondary user connection, resource matched to meet high energy efficiency.
7. the cognition wireless network that a kind of service-oriented according to claim 5 moderately services transmits learning method, special
Sign is: secondary user's connection status includes frequency range MkThe condition of Packet Error Ratio, the secondary user that priority j undergoes i-th to interrupt
In frequency range MkOn arrival rate and service time, given for one secondary user connection, the network before i+1 time is interrupted
State is expressed as
Wherein,For frequency range MkThe condition of Packet Error Ratio,WithRespectively type priority grade j undergoes time that i-th is interrupted
User is in frequency range MkOn arrival rate and service time；
Secondary user's connection activity refers to: when secondary user's disconnecting, needing to select to stay in current frequency range, or exchange to another
In one available frequency band, enableIt is secondary user in state sj,iAnd candidate active when i+1 time interruption,For Frequency Band Selection parameter, it is used to determine selection frequency range MkIt is current as transmission band and the probability when i+1 time is interrupted
When frequency spectrum handover delay being hoped to be more than the time delay off period, data packet can lose；
The secondary user connects return and refers to: for switching, the prediction that a movable return r is defined as multimedia transmission is end-to-end
Efficiency；It is expected that end-to-end efficiency consists of two parts, i.e., due to the Packet Error Ratio of channel condition and when frequency spectrum switching expectation time delay is super
The packet loss for spending the time delay off period, by this Reward Program, mandatory learning is used to maximize end-to-end efficiency, and balancing spectral
Switch caused by time delay, the expectation time delay of applied business high priority, will not by the secondary customer impact of other low priorities, but
Be if secondary user belongs to low priority, can consider using the influence of high-priority traffic as end-to-end performance criteria, and according to
Queuing model priority-based；
The on-line study of secondary user's connection refers to: target of the intelligent Agent when i+1 time is interrupted is that searching one is optimal
Activity, in current policies MOPT (sj,i,aj,i) under maximize end-to-end efficiency.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610046073.3A CN105636062B (en) | 2016-01-25 | 2016-01-25 | A kind of cognition wireless network that service-oriented moderately services transmission learning method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610046073.3A CN105636062B (en) | 2016-01-25 | 2016-01-25 | A kind of cognition wireless network that service-oriented moderately services transmission learning method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN105636062A CN105636062A (en) | 2016-06-01 |
| CN105636062B true CN105636062B (en) | 2019-02-26 |
Family
ID=56050446
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610046073.3A Expired - Fee Related CN105636062B (en) | 2016-01-25 | 2016-01-25 | A kind of cognition wireless network that service-oriented moderately services transmission learning method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105636062B (en) |
Families Citing this family (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108347734B (en) * | 2018-01-04 | 2021-11-23 | 西安电子科技大学 | Time-equalization-based resource allocation method in cognitive wireless communication system |
| US11800408B2 (en) | 2018-08-14 | 2023-10-24 | Telefonaktiebolaget Lm Ericsson (Publ) | Method for advance notification of changes to network QoS capabilities |
| CN109219071B (en) * | 2018-11-05 | 2021-09-10 | 重庆邮电大学 | Vertical switching method based on service classification in heterogeneous wireless network |
| CN110121206B (en) * | 2019-05-15 | 2021-06-08 | 重庆邮电大学 | Multi-hop wireless awakening method based on limited queue length queuing model |
| CN110493873B (en) * | 2019-07-02 | 2023-06-30 | 全球能源互联网研究院有限公司 | Wireless private network spectrum allocation optimization method and device suitable for power service |
| US20220358335A1 (en) * | 2019-08-30 | 2022-11-10 | Telefonaktiebolaget Lm Ericsson (Publ) | User plane selection using reinforcement learning |
| CN117135693B (en) * | 2023-10-27 | 2024-01-23 | 四川长虹新网科技有限责任公司 | Real-time service distribution method based on federal learning under multi-AP environment |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101494593A (en) * | 2008-01-25 | 2009-07-29 | 上海摩波彼克半导体有限公司 | Method for implementing optimization unite connection allowance and route control of wireless mesh network |
| CN103618674A (en) * | 2013-10-24 | 2014-03-05 | 上海交通大学 | A united packet scheduling and channel allocation routing method based on an adaptive service model |
| CN104486800A (en) * | 2014-12-09 | 2015-04-01 | 上海交通大学 | Moderate service-based adaptive cooperative transmission method and system in wireless network |
-
2016
- 2016-01-25 CN CN201610046073.3A patent/CN105636062B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101494593A (en) * | 2008-01-25 | 2009-07-29 | 上海摩波彼克半导体有限公司 | Method for implementing optimization unite connection allowance and route control of wireless mesh network |
| CN103618674A (en) * | 2013-10-24 | 2014-03-05 | 上海交通大学 | A united packet scheduling and channel allocation routing method based on an adaptive service model |
| CN104486800A (en) * | 2014-12-09 | 2015-04-01 | 上海交通大学 | Moderate service-based adaptive cooperative transmission method and system in wireless network |
Non-Patent Citations (2)
| Title |
|---|
| 移动无线传感器网络QoS机制研究;余华平;《长江大学学报》;20081215;全文 |
| 认知无线电资源自主分配和协同验证的研究;秦航;《长江大学学报》;20100318;全文 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105636062A (en) | 2016-06-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105636062B (en) | A kind of cognition wireless network that service-oriented moderately services transmission learning method | |
| Chen et al. | Computation offloading in beyond 5G networks: A distributed learning framework and applications | |
| Sheng et al. | Energy efficiency and delay tradeoff in device-to-device communications underlaying cellular networks | |
| Saleem et al. | SMART: A SpectruM-Aware ClusteR-based rouTing scheme for distributed cognitive radio networks | |
| Yau et al. | Application of reinforcement learning to wireless sensor networks: models and algorithms | |
| CN103108371B (en) | Single antenna multichannel cognition wireless sensor network route control method | |
| Jaber et al. | A distributed SON-based user-centric backhaul provisioning scheme | |
| Hossain et al. | Multi-objective Harris hawks optimization algorithm based 2-Hop routing algorithm for CR-VANET | |
| Niyato et al. | Optimization of the mobile router and traffic sources in vehicular delay-tolerant network | |
| Bao et al. | A QoE-maximization-based vertical handover scheme for VLC heterogeneous networks | |
| Seferoglu et al. | Diff-Max: Separation of routing and scheduling in backpressure-based wireless networks | |
| Ioannou et al. | A novel Distributed AI framework with ML for D2D communication in 5G/6G networks | |
| Amraoui et al. | Intelligent wireless communication system using cognitive radio | |
| Gao et al. | Freshness-aware age optimization for multipath TCP over software defined networks | |
| Kaur et al. | Intelligent spectrum management based on reinforcement learning schemes in cooperative cognitive radio networks | |
| Wang et al. | Admission control and channel allocation for supporting real-time applications in cognitive radio networks | |
| Wang et al. | Multichannel opportunistic access by overhearing primary ARQ messages | |
| Safdar Malik et al. | Reinforcement learning-based routing protocol to minimize channel switching and interference for cognitive radio networks | |
| EP4060948A1 (en) | Method for calculating tsn capabilities of a wireless bridge | |
| Usaha et al. | A reinforcement learning ticket-based probing path discovery scheme for MANETs | |
| Lopez-Martinez et al. | A superprocess with upper confidence bounds for cooperative spectrum sharing | |
| Al-Rawi et al. | Route selection for minimizing interference to primary users in Cognitive Radio Networks: A Reinforcement Learning approach | |
| Bello et al. | An Intelligent Hybrid Radio Access Technology Selection Algorithm for 5G-Satellite Network | |
| Shamani et al. | MPTCP energy enhancement paradox: A Q-learning approach | |
| Singh et al. | Ultra-dense low data rate (UDLD) communication in the THz |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190226 |